1. Lecture VIII: Software Architecture
CS 4593 Cloud-Oriented Big Data and Software Engineering

2. Today’s class Software Architecture 2 What is software architecture
Why software architecture
Software architecture styles 2

3. What is software architecture
Software architecture is the structure of a software system – like the blue prints in building architecture
Software components
Details (data structure and algorithms) hidden
Relationships among the components
Relations can be vague
Data flows
Control flows
Dependencies 3

4. Why software architecture?
Software becomes larger and larger
500 lines of code: just write it
5,000 lines of code: maybe class diagram (data flow diagram for non OO approaches)
500,000 lines of code: software architecture 4

5. A simple class diagram 5

6. A class diagram with 300 classes6

7. So we need higher level abstraction
Consider higher level concepts in architecture : components
For OO approach, components are usually a collection of classes
Consider the bookstore project:
For a small book store: books, shelves, clerks
But for barnes&noble: book storage, ads, finance, online stores, readers, bookstore is one component 7

8. Software Architectural Design
Usually used for large software projects and software frameworks
Early phase design
Important for task assignment
At a higher level than class diagrams
Components and their relationship
Interface can be vague at first 8

9. Software Architectural Design
Usually done by a small number of high-level people
It can be both technical and artistic
Experience and creativity can be major factors
It is not something you can learn well from the book
But there is something to learn about it 9

10. Software architecture styles
Something like architecture styles for buildings
Certain way to organize components and their relationships
Can be reused in different software
Use different architecture styles according to the usages of software 10

11. Architecture styles
Stadium 11

12. Architecture styles
Stations and airports 12

13. Architecture styles Skyscrapers Why no skyscrapers
looking like a stadium? 13

14. Some times people try new styles…14

15. Why Software architecture styles
Easier to do high level design
Barnes&Noble example
Consider an architecture like this: 15

16. Why Software architecture styles
Actually, most such information systems has an architecture like this, why?
Application
Service
Data 16

17. Why Software architecture styles
Reduce high level design risks
People may have tried different styles and found this best (It is not as obvious as for buildings)
The drawbacks of commonly used architectures are well studied
You may try new ones:
But not until you
are a really good
architect

18. Popular architecture styles
Pipe and Filter
Layered
Repository

19. Pipe and Filter 19 A defined series of independent computations
Performed for data transformation
A component reads streams of data on its inputs and produces streams of data on its outputs 19

20. Pipe and Filter 20

21. Pipe and Filter: Structure
Components: filter
Perform changes on the data input and generate output
Computation can be done with part of input, so that it can start before accepting all inputs
Connectors: pipe
Simply data transfer between filters

22. Pipe and Filter: Characteristic
Filters do not share state with other filters
Filters do not know the identity of their upstream or downstream filters
The correctness of the output of a pipe and filter network shall not depend on the order in which their filters perform their incremental processing 22

23. Pipe and Filter Examples
Unix Shell Scripts: provides a notation for connecting Unix processes via pipes.
e.g., cat file | grep err | wc
Compilers: the phases in the pipeline include:
Lexical analysis
Parsing
Semantic analysis
Code generation 23

24. Pipe and Filter - Advantages
Easy to understand the overall input/output behavior of a system as a simple composition of the behaviors of the individual filters
They support reuse, since any two filters can be hooked together, provided they agree on the data that is being transmitted between them 24

25. Pipe and Filter - Advantages
Systems can be easily maintained and enhanced, since new filters can be added to existing systems and old filters can be replaced by improved ones
The naturally support concurrent execution: map-reduce is actually pipe and filter architecture 25

26. Pipe and Filter - Disadvantages
Not good for handling reactive systems, because of their transformational character
Excessive parsing and un-parsing leads to loss of performance and increased complexity in writing the filters themselves 26

27. Layered Style 27 The system is divided to multiple layers
Each layer provides service to the layer above it and serves as a client to the layer below it
Each layer may have multiple components, these components usually do not interact much with each other 27

28. Layered Style 28

29. Layered Style: Structure
Components: are typically collections of procedures
Connectors: the layer margin, are typically procedure calls under restricted visibility
Layer Communication Rules:
Usually a component will talk only with the layer just beneath it
Only carefully selected procedures from the inner layers are made available (exported) to their adjacent outer layer 29

30. Layered Style Examples
Operating Systems
Unix
Windows
Android
…(almost any)
Distributed Information Systems
Web applications
Data/Logic/Appearance 30

31. Layered Style Advantages
Design: based on increasing levels of abstraction
Maintainability: changes to the function of one layer affects at most two other layers
Reuse: different implementations (with identical interfaces) of the same layer can be used interchangeably 31

32. Layered Style Disadvantages
Not all systems are easily structured in a layered fashion
Performance requirements may force the coupling of high-level functions to their lower-level implementations 32

33. Repository Style
Suitable for applications in which the central issue is establishing, augmenting, and maintaining a complex central body of information
Typically the information must be manipulated in a variety of ways
Often long-term persistence of information is required 33

34. Repository Style 34

35. Repository Style: Structure
Components:
A central data structure representing the correct state of the system
A collection of independent components that operate on the central data structure
Connectors:
Typically procedure calls or direct memory accesses 35

36. Repository Style: Characteristics
Components only interact with the repository
All components can access the repository
The repository controls and manages the access to its data 36

37. Repository Style Examples
Programming Environments
Graphical Editors
AI Knowledge Bases 37

38. Repository Style Advantages
Efficient way to store large amounts of data
Sharing model is published as the repository schema
Centralized management:
backup
security
concurrency control 38

39. Repository Style Disadvantages
Must agree on a data model a prior
Difficult to distribute data
Data evolution is expensive 39

40. Thanks! 40